Hi,
I know it is possible to parallelize process in the simulation stage. Is it possible to do the same when fitting parameters?
I have parallelized parameter optimization in the differetnialEquation Library, using blackboxoptim libray and :separable_nes method. I am not sure if something like this is available in PUMAS or how would you setup parallel fitting.
Could you let me kow if its possible and if so what’s the correct setup?
Thanks,
A
Parallel fitting is currently not supported but it’s something that we expect to have available pretty soon. At least a threaded version. Would that work for you or are you looking for distributed parallelism?
distributed is ideal, but threaded is also great-
Although I had not dabbled much in parallel programming, I have used the distribution and threading to increase optimization speed and both have helped a lot, with somewhat better results with distributed.
Thanks for the update.
A